Recently, proportional mean residual life model has received a lot of attention after the importance of the model was discussed by Zahedi [17]. In this paper, we define dynamic proportional mean residual life model and study its properties for different aging classes. The closure of this model under different stochastic orders is also discussed. Many examples are presented to illustrate different properties of the model.for t ∈ [0, ∞) such thatF X (t) > 0. It is to be mentioned here that if the support of X is [a, b], thenF X (t) = 0 for all t > b. In such a case, Eq. (1.1) does not define m X (t) for t > b.But it is customary, in such cases, to define m X (t) = 0 for t > b; see, for instance, Shaked and Shanthikumar [16, pp. 81-82]. Thus, m X (t) 0 for all t 0.Failure time data have been modeled by using Cox's proportional hazards (PH) model; see Cox [2]. Based on the assumption of proportional mean residual life (PMRL) functions, the concept of Proportional Mean Residual Life Model, parallel to Cox's PH model, was introduced by Zahedi [17] (see also Lam [7], Oakes [14], Oakes and Dasu [15]) as(1.2)The following interpretation for the PMRL model has been given recently by Nanda, Bhattacharjee, and Balakrishnan [12]. Let a series system be formed with k components, of which one component has lifetime distribution F and the other k − 1 components have i.i.d. life distribution, which is the equilibrium distribution corresponding to F . An equilibrium distribution is obtained as a limiting distribution of a renewal process. Then the MRL function of the system so formed and the MRL function corresponding to F are proportional with constant of proportionality c = 1/k. This paper deals with the properties of the PMRL model when the constant of proportionality depends on time. Gupta, Gupta, and Gupta [4] proposed the Proportional Reversed Hazard (PRH) Model to analyze failure time data. Gupta and Wu [6] have studied some properties of the PRH model. Mi [9] has shown that if a component has a bathtub-shaped failure rate function, then MRL is unimodal, but the converse does not hold. Later, Ghai and Mi [3] developed sufficient conditions for the unimodal MRL to imply that the failure rate function has a bathtub shape. The PMRL model and its implications have been discussed by Gupta and Kirmani [5]. Nanda, Bhattacharjee, and Alam [10] have discussed some properties of the PMRL model in the context of reliability theory. The PMRL model has been extended to a regression model with explanatory variables by Magulury and Zhang [8]. If we replace c in Eq. (1.2) by some non-negative function of t, say c(t), then the corresponding equation becomes